Experimental Thermal and Fluid Science 73 (2016) 87–93
Contents lists available at ScienceDirect
Experimental Thermal and Fluid Science
journal homepage: www.elsevier.com/locate/etfs
A burner to emulate condensed phase fuels
Y. Zhang a, M. Kim a, P.B. Sunderland a, J.G. Quintiere a,⇑, J. de Ris b,1
a
b
Dept. of Fire Protection Engineering, University of Maryland, College Park, USA
FM Global, USA
a r t i c l e
i n f o
Article history:
Received 22 September 2015
Accepted 22 September 2015
Available online 30 September 2015
Keywords:
Burner
Emulating
Pool fires
a b s t r a c t
A gas-fueled burner with heat flux gages embedded in its porous surface is used to emulate condensed
fuel flames. The measured heat flux, the flow rate of the fuel/inert mixture, and the burner surface
temperature allow the emulation of the burning characteristics of condensed fuels. The burner is named
the Burning Rate Emulator (BRE). It can burn a gaseous fuel at an effective heat of gasification matching
the actual heat of gasification of condensed-phase fuels. It also can match other characteristics of the
condensed-phase fuel by careful selection of certain properties of the gaseous fuel. These properties
are the heat of combustion, the effective heat of gasification, the surface temperature, and the laminar
smoke point. The BRE is shown to reasonably emulate steady burning of methanol, heptane,
polyoxymethylene (POM) and polymethylmethacrylate (PMMA) burning in 50 mm diameter pools. It also
can be used to emulate ignition and extinction. The results can be used to predict behavior at other
conditions, such as burning with external radiant heating. The BRE can be extended to emulate steady
burning under diverse conditions. The plausibility of the BRE is demonstrated and its limitations and difficulties are discussed. In particular, the difficulty of dealing with the actual surface heat flux distribution
is examined. In general, the paper intends to demonstrate the attributes of a BRE.
Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction
This study seeks to establish the burning conditions for
condensed fuels using a gaseous burner. The burner conditions
represent those of steady burning with the heat of gasification of
the material as the principal fuel property. In addition three other
properties are put forth to complete an emulation. They include
the heat of combustion needed to control flame extent,
re-radiation heat flux governing surface heat loss, and flame radiation represented by the laminar smoke point. These properties
are hypothesized to establish the physical identity of the fuel its
chemistry is not directly considered.
We seek to define the burning rate in terms of these four
properties by using an emulator having a controlled gaseous fuel
supply. The chemical nature of the gas is not modeled in respect
to the real fuel’s chemical composition. The Burning Rate Emulator
(BRE) can be operated to simulate a condensed-phase fuel in
steady burning. The results will depend on size, orientation, and,
environmental conditions; however, we will examine primarily
burning in air.
⇑ Corresponding author.
1
E-mail address: [email protected] (J.G. Quintiere).
Retired.
http://dx.doi.org/10.1016/j.expthermflusci.2015.09.025
0894-1777/Ó 2015 Elsevier Inc. All rights reserved.
The theory of steady burning for an evaporating condensed fuel
is considered as the basis of the BRE. Although steady burning is
not practical for many condensed fuels, the BRE can still give valuable insight on average. In other words, as wood would have a nonsteady burning signature due to charring, its peak or overall average burning rate can be represented by an appropriate (usually a
high) value for the heat of gasification. Steady burning (rate per
unit area or burning flux) can be formulated in terms of a heat of
gasification (L) as
_ 00 ¼
m
q_ 00f q_ 00rr þ q_ 00e
L
;
ð1Þ
where the quantities are defined in the nomenclature. The heat of
gasification for liquids is a thermodynamic property defined as
L ¼ hv ap þ cp ðT v T 1 Þ;
ð2Þ
where hvap is the heat of vaporization, cp is the specific heat, Tv is the
surface vaporization temperature, and T1 is the ambient temperature. The heat of gasification of polymers will involve more phases
and transitions. As stated even charring materials, in time-average
burning, can be couched in terms of a relatively high value of L.
The flame heat flux is controlled by convective and radiative
components, and the surface re-radiation heat flux by the temperature of the vaporizing surface. In the case of the BRE, it is the
88
Y. Zhang et al. / Experimental Thermal and Fluid Science 73 (2016) 87–93
Nomenclature
B
cp
hc
Dhc
hvap
_ 00
m
L
q_ 00e
B ½Y ox Dhc =r cp ðT v T 1 Þ=L (dimensionless)
specific heat of gas (J/g K)
convective heat transfer coefficient (W/m2 K)
heat of combustion (kJ/g)
heat of vaporization (kJ/g)
burning rate (g/m2 s)
heat of gasification (kJ/g)
incident external radiative heat flux (kW/m2)
burner surface temperature that gives the surface re-radiation.
Heat flux gages in the burner surface record the flame heat flux,
which includes both incident radiation and convection to the
surface. With the measured gas flow rate of the burner, values of
L can be determined from Eq. (1). While such values might seem
simplistic, an effective L, which includes external heating, can
define the enhancement conditions needed for that L to burn. By
varying the gaseous fuel–inert mixture, the heat of combustion
and the radiation character (specifically the laminar smoke point)
can be varied. For a given configuration, a wide range of burning
conditions can be readily established with the emulator (BRE).
But the development of the required net heat flux of Eq. (1) can
present a challenge, and the control of a burner’s surface temperature can be difficult. While the dilution of a gaseous fuel with an
inert gas might match the heat of combustion and the smoke point
of the emulated fuel, complete matching may only be approximately achieved.
The use of a burner to emulate combustion of solids and liquids
has been used previously examined. Orloff and de Ris [1], Kim et al.
[2], and de Ris et al. [3] pioneered the use of sintered metal burners
for studying the steady burning of a planar condensed phase. Their
burner used water cooling to obtain the needed heat flux, and thus
the surface temperature was low. The water cooling also led to
long equilibrium times. They examined mainly convective burning
through the Spalding B number. For a given ambient condition, the
B number is principally a function of the heat of gasification, L. The
burning rate per unit area in purely diffusive or convective burning
is also principally a function of the B number. Flame radiation and
surface re-radiation disturb this simple dependence. However, the
relatively simple dependence could be emulated by the de Ris burner, and is shown to follow laminar pure convective flame theory
[1–3].
The use of a burner is imperfect, as it generally maintains a uniform velocity over its surface. Boundary layer or pool burning, even
for pure convection, will have a distribution of heat flux over the
surface, and thus a variable surface velocity. The good agreement
with theory suggests that the fuel velocity at the burner face
quickly equilibrates to proper diffusional flows in the flame.
Bustamante et al. [4] presented another validation of burner
emulation by comparing the flame standoff distance in the laminar
region for burning of inclined flat plates. Results showed the
similarity in the flame shape for a flat surface oriented at various
angles. Even the onset of turbulent unsteady flow was approximately matched.
2. Experimental design and testing
Inspired by the de Ris burner, a BRE burner was designed and
constructed. Its size was selected to replicate small pool fires. This
BRE burner has a face diameter of 50 mm. Its internal features
allow for a mixing plenum for the incoming fuel stream, an array
of glass beads to provide uniformity to the flow, and a top brass
q_ 00f
q_ 00rr
r
Tv
T1
Yox
incident flame heat flux (kW/m2)
surface radiative loss heat flux to ambient (kW/m2)
stoichiometric mass oxygen to fuel ratio (g/g)
vaporization temperature (K)
ambient temperature (K)
ambient oxygen mass fraction (g/g)
plate with uniform holes having a high overall porosity. Two heat
flux sensors are used anticipating heat flux variation over the
surface. They are needed to compute the heat of gasification.
Two thermocouples on the operating face record the surface temperature so that the re-radiation heat flux can be computed. Fig. 1
shows the BRE burner with two heat flux sensors, one at the center
and the other between the center and edge. The ‘‘edge” sensor is at
a radius of 3.2 mm. The sensors are 1/8-in. diameter water-cooled
Medtherm thermopile devices, operated at about 65 °C to avoid a
condensation error. The required heat fluxes needed in Eq. (1)
are computed from these temperatures and heat flux measurements. A distribution needs to be postulated to give the integrated
average radial heat flux. Also corrections are applied to address the
differences in the sensor and plate temperatures. These corrections
are generally small, and details will not presented here.
A series of tests were performed to assess the burner’s ability to
emulate particular condensed-fuel combustion. The procedure was
to measure a steady burning rate for the condensed fuel, then
select a gaseous fuel mixture with this same flow rate to best
match (1) the heat of combustion, (2) the heat of gasification, (3)
the surface temperature, and (4) the laminar smoke point of the
original condensed-phase fuel. The heat of combustion and the
smoke point could be matched as close as practical by selecting a
mixture of pure gaseous fuel and an inert diluent; nitrogen was
used.
Several condensed-phase fuels were selected for study: two
liquids and two solids. Two of the fuels burn with negligible soot.
The four properties were matched as closely as practical, and the
burning rates, flame size, character, and color were compared.
The burning rate of the condensed fuel is determined by a loadcell measurement of a 50 mm pool fire, and used to fix the BRE flow
rate. The liquids were burned in a 50 mm diameter glass vessel.
Although the liquid fuel was not refilled during the burning, we
have reached the stable burning rate. Literature values are established for the four fuel properties as listed above [5–7]. Two of
the four properties – heat of combustion and laminar smoke point
– are best matched by selecting the proper gaseous fuel and nitrogen mixture. The other two properties – heat of gasification and
surface temperature – are found from the BRE measurements. As
the burner surface temperature is not controlled, this matching
depends on the burner thermal properties. As long as the surface
temperatures are not too out of line, this is not considered a serious
defect to the emulation.
The four fuels selected are methanol, heptane, PMMA and POM.
The liquids easily ignite and quickly establish steady burning. The
solids need to be encouraged to burn, but eventually reach steady
conditions.
Methanol was examined first. The mass loss rate of methanol
was measured and the flame was photographed. To match the
methanol with the BRE we would need to use a gaseous fuel with
the same flow rate. To match the flame height, the fuel must have
the same heat of combustion. In addition, consideration of the soot
89
100 mm
Y. Zhang et al. / Experimental Thermal and Fluid Science 73 (2016) 87–93
tendency of methanol led to the choice of methane as the gaseous
fuel for the BRE. Nitrogen was added to the fuel stream to match
the heat of combustion of methanol, 19.1 kJ/g [6]. The gas burner
BRE used a fuel mixture of 52% (volume) methane with 48% nitrogen. The flow rate of the mixture was 24 cc/s to match the burning
rate of the methanol at 11 g/m2 s. If the BRE concept is correct, the
two flames should be nearly identical in appearance and in flame
heat flux distribution. The measured heat fluxes in the BRE are
5.5 and 21.3 kW/m2 for the heat flux gages located at the center
and at 1.5 cm from the center, respectively. A weighted average
over their area segments gives an average value of 15.6 kW/m2.
Corresponding surface temperature measurements yield an average temperature of 160 °C (141–170 °C). Taking an emissivity of
unity for the BRE oxidized porous brass surface, the re-radiative
heat flux is approximately 2 kW/m2. Consequently, the heat of
gasification associated with the BRE as computed from Eq. (1) is
L = (15.6–2 kW/m2)/(11 g/m2 s) = 1.24 kJ/g. The literature value
for methanol is 1.20 kJ/g [7]. In addition to small approximations
in the calculations, there is a difference of the BRE surface, as it
is hotter (160 °C) than the boiling point of methanol at 64 °C. This
temperature discrepancy is beyond the control of the BRE, as its
surface temperature is a result of the flame and burner heat
transfer characteristics.
A visual confirmation of the BRE to emulate the methanol pool
fire at 50 mm diameter is shown in Fig. 2a and b, in which the
flames are compared when the slightly oscillating images are
similar. The color, size, and oscillatory character are similar.
We then sought to evaluate whether the burner could also emulate other burning conditions. We considered a 50-mm-diameter
burning pool of heptane. Here the flame would be yellow from
soot. The choice of the gaseous fuel in the BRE was again made
by matching the heat of combustion and the laminar smoke point
for heptane. Ethylene was selected be used as the burner gas to
simulate heptane. The heat of combustion of ethylene is close to
that of heptane (41.5 kJ/g compared to 41.2 kJ/g) [6], as is the
laminar smoke point (120 mm compared to 139 mm) [8]. A visual
confirmation for the nature of these turbulent heptane flames is
shown in Fig. 3.
Two solids were then examined: PMMA and POM. The latter
burns with negligible soot. A propylene and nitrogen mixture
(50% in mole fraction) was used to emulate PMMA. Steady burning
of the PMMA was achieved by igniting it in an inverted orientation
(facing down) and allowing the flame to stabilize, after which the
sample was oriented as a pool fire where it sustained steady
Methanol Pool
BRE
Fig. 2. Comparison of 50 mm diameter methanol flame reproduced by the BRE
burner.
200 mm
Fig. 1. BRE burner with 50 mm diameter surface.
Heptane
BRE
Fig. 3. Comparison of 50 mm diameter heptane flame reproduced by the BRE
burner.
burning. A visual confirmation for the comparison of PMMA and
BRE is shown in Fig. 4.
The results for POM are shown in Fig. 5. As with the PMMA, the
POM was cut to 50 mm diameter disk. Before placed on a scale, it
was first heated with pilot flame until the flame covered the whole
burning surface. The POM burnt steadily and that value was
Y. Zhang et al. / Experimental Thermal and Fluid Science 73 (2016) 87–93
fires of 50 mm diameter is fraught with issues. It is known that for
liquid pools edge heat losses to the container, and the container lip
influence the burning rate. Also refilling the liquid to maintain a
fixed level has an effect. Many investigators have tried to minimize
these issues. Yet, as seen for example in Fig. 6, the burning rates
presented in the literature for methanol, including ours, have
differences [8–11].
In addition, the BRE uses two point heat flux measurements to
obtain the average net heat flux required by Eq. (1). To be accurate,
these points must be incorporated into a proper distribution across
the radius of the burner. Theory or experimental results can guide
this. Unfortunately there has been little done in this area. The work
of Akita and Yumoto [8] shed some light on the heat flux distribution. They measured the burning rate in small pool fires over three
concentric regions of the pool. From this the heat flux can be
deduced, and the results are shown in Fig. 7. It is shown that the
related heat flux distribution can be exponential over the radius.
(Note the heat flux is proportional to the supply velocity, Eq. (1)).
This sharp increase in heat flux to the edge can be appreciated
by examining the flame shape for ethylene as shown in Fig. 8.
Two features should be noted, (1) the closeness of the flame at
the edge giving high convection heat transfer, and (2) the color
of the body of the flame indicating the presence of radiation.
In processing the data for the four fuel emulations, the heat flux
was assumed to step change for the two sensors. This is shown by
100 mm
90
PMMA
BRE
Fig. 4. Comparison of 50 mm diameter PMMA flame reproduced by the BRE burner.
q_ 00av g ¼
q_ 00av g
1
p2:52
Z
0
2:5
q_ 00 2prdr ¼ q_ 00r¼0 cm
2
2
1:5
1:5
þ q_ 00r¼1:5 cm 1 2:5
2:5
¼ 0:36q_ 00r¼0 cm þ 0:64q_ 00r¼1:5 cm
20 mm
ð3Þ
POM
BRE
Fig. 5. Comparison of 50 mm diameter POM flame reproduced by the BRE burner.
matched by the BRE. The color and nature of the BRE flame is well
matched by the POM mixture. The relevant properties are given in
Table 1 along with those of the other fuels. The BRE cannot match
the surface temperature, but more importantly the heat of gasification is nearly matched in all cases. These examples confirm the
implicit hypothesis of the four properties being sufficient to
establish the same burning condition with the BRE.
3. Difficulties
It should be noted that the demonstration of the BRE to emulate
real fuel fires is not without difficulties. The burning of small pool
Obviously the nature of the distribution is key to interpreting
the two sensors. We are continuing to work this issue and are currently in the process of measuring a more complete distribution
with an array of sensors. Just to show the differences that can be
developed, we compare the individual sensors, with the distribution in Eq. (3), and the average heat flux measured by a perforated
relatively thick copper exit plate for the burner. This top plate is
used as a calorimeter, and its heat flux is deduced by a heat transfer analysis. This has been incorporated into a new burner design,
and it is still under development. As we assure its accuracy, we
would now have a way to derive the average as well as two point
measurements for the heat flux. A preliminary result is shown in
Fig. 9. The figure shows that the heat flux tends to become more
uniform as the burning rate increases. The average heat flux
decreases less with increasing mass transfer. This suggests that
the peak heat flux occurs at larger radii with increasing mass flux.
This is expected as the flame is pushed farther away from the surface. The results show the top plate calorimeter is about 15% higher
than that using Eq. (3).
These difficulties suggest that the very good results for the
emulation in Table 1 could be fortuitous. However, it should be
realized that the BRE is matching the burning rate obtained in
Table 1
Literature [6–8] and measured properties that determine the burning conditions.
_ 00 (g/m2 s)
m
Dhc (kJ/g)
SP (mm)
Ts (°C)
L (kJ/g)
Methanol
Pool
BRE (XCH4 = 52%, XN2 = 48%)
11
11
19
19
1
1
64
160
1.2
1.24
Heptane
Pool
BRE (C2H4)
15
15
41.2
41.5
139
120
98
211
0.48
0.51
PMMA
Pool
BRE (XC3H6 = 50%, XN2 = 50%)
6
6
24.2
24.3
105
117
390
312
1.6
1.8
POM
Pool
BRE (XCH4 = 41%, XN2 = 59%)
9
9
14.4
14.1
1
1
420
167
2.4
2.1
91
Y. Zhang et al. / Experimental Thermal and Fluid Science 73 (2016) 87–93
burning rate (g/m2-s)
Methanol
Corlett (reill)
J. Gore (reill)
Yi (not reilled)
20
H. Hayasaka (not
reilled)
K. Akita (reill)
Fig. 8. Ethylene flame on the BRE 50 mm burner.
10
1
10
100
Pool Diameter (cm)
Fig. 6. Burning rates of methanol pool fires of small diameter [8,9].
our experiments though they might not be the most accurate
steady state results. Also the two-point integration to get the average heat flux might be satisfactory, with a 15% possible error, but
our evaluation is still in progress. Yet the color and height of the
flames add to the credibility of the BRE. Moreover, past work supports its performance [1–4].
4. Other applications for the BRE
Having demonstrated the BRE’s capability to emulate the combustion of various fuels in normal gravity, extensions of its use can
be considered. It provides a very controllable way to study steady
burning. Its configuration and size can be varied. It can be used in
many different environments.
As in the studies by de Ris et al. [1–3], the burner can be operated over a range of L or B values. This is particularly applicable to
purely convective burning. We demonstrated this for the 50 mm
BRE using a mixture of methane and nitrogen to generate data over
a range of B values. The results correlated with ln(1 + B) giving a
convective heat transfer coefficient of 10.5 W/m2 K are plotted in
Fig. 10. Along with the data of the BRE with diluted methane, several real condensed fuels were added. As methane tends to have a
relatively low radiation loss, as well as methanol, the PMMA and
heptane having higher radiative losses could explain their variation. The higher the radiation loss, the lower the flame temperature
and for a small diameter pool where convection dominates, the
burning rate is apt to be lower as shown on the graph.
Fig. 9. Heat flux to the 50 mm BRE as a function of flow rate of pure methane.
Heptane
Methanol
POM
PMMA
Fig. 10. BRE steady burning for CH4–N2 mixtures at 50 mm and ‘‘real” solid/liquids.
Fig. 7. Methanol supply rates over three concentric section of a 10.66 cm diameter
pool fire (from [8]).
Mass-loss flux at ignition and extinction limits are important
properties of fires. Yet their evaluation for condensed fuels is difficult to capture because of their highly transient nature. Materials
experience a sudden jump in mass-loss rate when ignited or when
they suddenly extinguish. Such transients are difficult to accurately measure. By using the BRE, the mass-loss rate can be clearly
identified by increasing the flow rate of fuel gradually until the
sustained ignition is observed, or decreasing it to extinction after
a flame is sustained.
92
Y. Zhang et al. / Experimental Thermal and Fluid Science 73 (2016) 87–93
considered to be complete. In contrast to the steady domain of
the BRE results, data for solids and liquid fuels are included. The
charring materials lie outside of the ‘‘flammability” domain, and
are well known not to burn without the support of external heating
or increased ambient oxygen. Liquid fuels and non-charring solids
do fall within the flammability domain as might be expected. A
study along these lines is ongoing, and it will be challenging on
how to present it, verify it, and explain it.
250
Methane
Propylene
200
ṁ"Δ hc kW/m2
Propane
Ethylene
150
Real Material (Lyon et al.)
Theory
5. Conclusions
100
50
0
0
10
20
30
40
50
60
Δhc kJ/g
Fig. 11. Extinction conditions for various solid fuels and data from the BRE for
several gaseous fuels [12].
25
CH4/N2
Liquid
Solid
Charring Solid
20
L kJ/g
15
10
5
0
0
10
20
30
40
50
60
Δhc kJ/g
Fig. 12. Flammability map by BRE in normal gravity (arrows toward the right fix
the nitrogen flow rate and increase the methane flow rate from 0 to 2 slpm). The
boundary with the arrow to the left is close to the fire point data for the fuel
mixture.
Lyon and Quintiere [12] compiled extinction data for condensed
fuels and examined their dependency with the heat of combustion.
Those data are compared in Fig. 11 for four gaseous fuels mixed
with nitrogen in the BRE. As the fuels have differing laminar smoke
points, the trends differ for each and suggest an effect of radiation.
The theoretical curve is taken from the analysis in [12]. We are in
the process of further studying these limits.
The BRE can be used with various fuel mixtures to prescribe the
heat of combustion and laminar smoke point. Over a range of flow
rates where a steady flame is sustained, the BRE then establishes
the corresponding heat of gasification and burning temperature.
These burning points can be plotted in a multi-dimensional format
to display the range of conditions that support steady burning.
Fig. 12 is a sample of such a plot in two dimensions that shows
the steady state domain of the 50 mm BRE in air with coordinates
of heat of combustion and heat of gasification. Re-radiation
heat flux and laminar smoke point are ignored here, but must be
The BRE is capable of matching the burning rate and flame characteristics of a condensed-phase fuel by maintaining the heats of
combustion and gasification, burning surface temperature and
the laminar smoke point the same, as practicable. This has been
demonstrated for pool fires of methanol, heptane, PMMA, and
POM. However even with the success of these demonstrations,
questions exist about the interpretation of the heat flux distribution, and on the accuracy of the burning rate measured for the
comparable pool fires. No good data or any theory exists for pool
fires, particularly in the range of 25–100 mm in diameter. Nevertheless the BRE concept allows for a more careful vehicle for the
study of these features over the difficulties with burning
condensed-phase fuels. We are engaged in work to measure the
average heat flux to the burner by means of a top-plate calorimeter, in addition to point measurements. The ability to compute an
accurate heat of gasification for the BRE depends on obtaining this
average heat flux determination.
Extinction and ignition conditions for the burning of
condensed-phase fuels can be studied much more easily and accurately with the BRE. While standard tests exist for measuring the
flash and fire-point of liquid fuels, they do not yield the mass flux
for these conditions. Modelers are currently using such critical
mass flux values to predict the ignition of solids. The BRE is capable
of giving this value in a precise and well-defined manner. Definitions of the flash, fire and extinction points can be sharp with
the BRE. In our ongoing studies, the BRE demonstrates that the fire
and extinction points do not necessarily have full fuel diameter
flames at these transitions. For solids and liquid fuel fuels such
transitions are very transient and rapidly move from full flame to
no flame, or vice versa. The BRE allows a steady snapshot study
of these transitions. Work is ongoing to elucidate and measure
these transition points, and hopefully to more fully demonstrate
their dependence on the heat of combustion.
The BRE offers a convenient way to study the prospect of steady
burning in various ambient conditions including microgravity. Of
course experiments in microgravity must be done in orbit to allow
sufficient time to reach valid end states. However studies with the
BRE in 5 s drop experiments have given interesting results.
Although the flames are still slowly evolving at the end of the
microgravity drop, a steady state convective theory was capable
for correlating the burning rate and final size of the flames over a
range of B numbers [13].
In conclusion, the concept of the BRE defined herein offers a
platform for the study of diffusion flames related to real
condensed-phase fuels. It offers advantages by allowing steady
conditions to prevail long enough to make detailed measurements
of variables of interest, and to better defined the limits of burning
at ignition and extinction.
Acknowledgment
This work was funded by NASA Grant NNX10AD98G for which
Paul Ferkul and Dennis Stocker have served as the technical
contact, and have provided important support.
Y. Zhang et al. / Experimental Thermal and Fluid Science 73 (2016) 87–93
References
[1]
[2]
[3]
[4]
L. Orloff, J. de Ris, Proc. Combust. Inst. 13 (1971) 979–992.
J.S. Kim, J. de Ris, W.F. Kroesser, Proc. Combust. Inst. 13 (1971) 949–961.
J. de Ris, L. Orloff, Proc. Combust. Inst. 15 (1975) 175–182.
M.J. Bustamante, K.T. Dotson, Y. Zhang, P.B. Sunderland, J.G. Quintiere, Laminar
Burning on Flat Wicks at Various Orientations, Spring Technical Meeting of the
Central States Section of the Combustion Institute, 2012.
[5] Y. Zhang, M.J. Bustamante, M.J. Gollner, P.B. Sunderland, J.G. Quintiere, J. Fire
Sci. 32 (2014) 52–71.
[6] J.G. Quintiere, Fundamentals of Fire Phenomena, John Wiley & Sons Ltd.,
Chichester, U.K., 2006.
[7] L. Li, P.B. Sunderland, Combust. Sci. Technol. 184 (2012) 829–841.
93
[8] K. Akita, T. Yumoto, Heat transfer in small pools and rates of burning of liquid
methanol, Proc. Combust. Inst. 10 (1965) 943–948.
[9] R.C. Corlett, T.M. Fu, Some recent experiments with pool fires, Pyrodynamics 1
(1966) 253–269.
[10] H. Hayasaka, Unsteady burning rates of small pool fires, Fire Saf. Sci. 5 (1997)
499–510.
[11] J. Gore, M. Klassen, A. Hamins, T. Kashiwagi, Fuel property effects on burning
rate and radiative transfer from liquid pool flames, Fire Saf. Sci. 3 (1991) 395–
404.
[12] R.E. Lyon, J.G. Quintiere, Combust. Flame 151 (2007) 551–559.
[13] Yi Zhang, Matt Kim, Haiqing Guo, Peter B. Sunderland, James G. Quintiere, John
deRis, Dennis P. Stocker, Emulation of condensed fuel flames with gases in
microgravity, Combust. Flame 162 (2015) 3449–3455.